Swansea castle has a long and distinguished history, but sadly the ruins of its castle are mostly gone.  The fortress may have been founded as early 1081 when King William I passed by on his way to St Davids, 65 miles west, following the main Roman coast road.  He certainly founded Cardiff castle, 40 miles to the east, on route.  Founding a castle here would have made sense, but that proves nothing.  If William did not do this then certainly a castle was standing here when it was attacked in 1116 during the reign of his son, Henry I (1100-35).  It was then held by Earl Henry Beaumont of Warwick (d.1119) as the caput of his lordship of Gower.  The fortress may have been threatened after 1136 when Gower was attacked, but it was only in 1192 that the castle suffered its major test, a 10 week long siege, followed by a battle in which the relieving forces shattered Rhys ap Gruffydd's besieging army.  King John helped strengthen the castle in 1210, after taking the castle back from William Braose.  In November 1215 the castle was taken by Llywelyn ab Iorwerth (d.1240) and Rhys Gryg (d.1234).  They returned it Braose's grandson, John (d.1232), who rebuilt it with Llywelyn's permission in 1221.  In 1257, after his victory over the Carmarthen army, Llywelyn ap Gruffydd (d.1282) repeated the feat of his grandfather in taking the castle.  In 1287 the town was sacked by Rhys ap Maredudd (d.1292), but the castle held out. 

For some unknown reason Bishop Henry Gower (1328-47), who obviously hailed from the Gower judging by his name, is thought to have built ‘the new castle' in the outer bailey of Swansea castle.  This is part of the structure that survives today.  The bishop also founded the Hospital of St David in Swansea in an act which seems similar to Bishop Bek's rebuilding of Llawhaden and its castle.  The attribution of the work to him is based on the fact that the exotic arcaded parapet on the one block is similar to those found at the bishop's palaces of St Davids and Lamphey.  The possibility that the lord of Swansea saw these and decided to copy them or hire their masons simply doesn't seem to have been put forward before.

Most knowledge of the old castle's form comes from observations of building work in the twentieth century.  This found a 34' deep ditch with a 9' wide flat bottom.  Obviously this was a water filled moat.  This seems to have formed the western part of a ringwork lying in the north-west portion of an early enclosure which was itself contained within the much later town walls.  This early enclosure is rectangular, about 350' east to west and 420' north to south and is bounded by the modern street system of Princess way, Caer St, The Strand and Welcome Lane/College St.  As such it is quite possible that the site was first fortified as a Roman fort.  Similar occurrences happened at Cardiff, Colwyn, Loughor and Painscastle.  Also Carew may be built on the site of a Roman fortlet.  In 1913, when buildings were built over the west side of the ringwork, a piece of curtain wall and a rectangular mural tower were briefly revealed.

The present remains of Swansea castle stand in the south-east corner of the later outer ward and consist of 2 separate blocks linked by a curtain wall.  They probably formed the south-east section of the outer ward.  At the north end is a roughly square tower that was used as the debtors' prison into the late eighteenth century.  This may have been similar to the inner ward tower foundations turned up in 1913 to the north of the remaining buildings.  Therefore the rectangular towers may represent the oldest surviving parts of the fortress.  The adjoining east curtain contains a stair entering a mural passage and giving access to the roof.  The lower section of this tower has distinct Sutton stone quoins and a single loop by the join with the south curtain.

The south block is much larger than the prison tower and appears to be built upon the outer bailey curtain walls on its exterior south side.  This terminates at the west end in a small turret, which is corbelled out on the third floor into a slightly larger structure.  This expansion seems to mark the original height of the turret.  The upper arcades of the block are clearly later than the curtain and lower portions of the turret.  Their purpose is obviously to provide the raised, but narrow wallwalk which had both parapet and parados.  Why a bishop might want to build this in the late fourteenth century is unknown.  They certainly would not have survived a serious assault with any artillery.

Externally in the south block there are several Sutton stone loops.  The one at the west end beyond the arcade has a normal stone lintel.  This leads to the possibility that all the loops were originally similar to this, although the loops in the merlons of the battlements above are all of Sutton stone and have ball oillets top and bottom.  The lower loops have no oillets.  The two loops in the south-east corner turret of the south-east tower also have 2 sighting slits with ball oillets to join the upper and lower ones as do the 3 loops in the basement hall and the 2 loops in the adjoining south-east tower.  The west facing loops in the north tower are all pointed windows.

Within the south block is a gabled first floor great hall with the lower 2 storeys of a D shaped tower half way along its internal north wall.  At the east end is a polygonal projecting tower.  The northern section may once have been another square tower, but it has an odd inversed triangular extension to the south.  At first floor level this entire structure forms one room, while there is also a smaller basement beneath.

Traces of a twin towered gatehouse were uncovered where Castle St meets College St centrally in the rectangular, outer ward north front.  The inner ward was apparently egg shaped and lay along the east side of the outer ward between Castle St and The Strand.  A layout such as this mimics the larger sites of Portchester and Pevensey, set within Roman forts.

Why not join me at other Lost Welsh Castles next Spring?  Please see the information on tours at Scholarly Sojourns.


Copyright©2019 Paul Martin Remfry